The present invention relates to monitoring framework, and more specifically to a distributed diagnosis architecture for remote monitoring, diagnosis, and maintenance of systems having multiple components, each component having sensors for monitoring and generating signals indicative of a characteristic of the component.
High rise buildings, nuclear power plants, manufacturing facilities, etc. generally employ a plurality of highly connected network enabled systems. For example, heating and air conditioning systems, fire prevention systems, alarm systems, security systems and many other systems and subsystems thereof perform various functions in these structures and facilities as well as in many other applications. Often, systems such as those identified above, perform critical functions and need to be closely monitored to ensure the efficient and safe operation of the facility or larger system. Systems such as those identified above, as well as others, could benefit from continuous monitoring for performance and maintenance reasons. Typically, systems such as these consist of embedded sensors and are network enabled allowing for the transmission of data to central (or remote) monitoring stations. Many of these systems are sensor rich, that is they employ numerous sensors for monitoring and generating signals indicative of a characteristic of the system, subsystem or component thereof.
Often, the structure, facility or larger systems lack the capability of monitoring the various systems and subsystems employed thereby. For example, various systems such as elevator systems, copy machines, automobiles, etc. employ many subsystems comprising sensors capable of transmitting signals for the purpose of performance monitoring and maintenance. However, many of these type of systems, or components thereof, as well as others, are usually not monitored and maintained remotely over a network.
Many such systems could benefit from a diagnosis server capable of processing data-streams from multiple components thereof by remote devices for the purpose of diagnosis and prognosis of the system or components thereof. Implementation of such a remote networked enabled device for monitoring the many systems and subsystems employed by facilities or large systems is significantly more cost effective than embedding devices capable of monitoring the performance of these systems within each of the structures or large systems.
In an increasingly connected world, it is not hard to imagine systems ranging from household appliances to sophisticated aircraft systems or large office complexes being continuously monitored or routinely connected to remote devices for services such as evaluation, diagnosis and maintenance of the systems and subsystems employed thereby.
Accordingly, a framework and architecture for remotely monitoring, diagnosing and maintaining a system having multiple components or subsystems thereof is disclosed.
The invention, in one aspect, discloses a diagnosis server and method that provides for continuous and remote monitoring and transmittal of sensor generated signals to networked logical reasoner applications which process the signals and provide diagnosis, failure detection, responsive signals and user output for the purpose of monitoring and maintaining systems of various complexity in real-time using applications remote from the system being monitored. The diagnosis server provides user-interactive reasoner applications for user controlled monitoring, diagnosing and maintaining of the system being monitored. Additionally, the present invention is designed to utilize prior art and future diagnostic applications for analysis and modeling of sensor signals providing efficient and accurate diagnosis and prognosis of a systems performance and identifying system failures and the potential therefor. The diagnosis server enhances preventive maintenance programs and reduces the number of unscheduled repairs for systems of various complexity in a wide range of industries.
The present invention network-based diagnosis server comprises an architecture such that the diagnostic reasoner applications and system specific knowledge are clearly separated. The latter includes database stored information comprising specifications and diagnostic information specific to the system being monitored as well as models of the system. Also, the diagnosis server provides signal processing and handling applications independent of the logical and user-interactive reasoner applications. These design features allow the diagnosis server to be truly distributed and readily adaptable to various different systems.
Based on the foregoing, it is a general object of the present invention to provide a method and apparatus for remotely monitoring, diagnosing and maintaining a system having multiple components or subsystem thereof.
One advantage of the invention is to reduce system down time by providing continuous and automatic analysis of the performance of individual components of the system being monitored and to identify and isolate faulty parts immediately upon failure, or before failure, enabling maintenance personnel to reduce or eliminate time spent troubleshooting failures in the system.
Another advantage of the invention is that it provides a distributed network-based diagnosis server for monitoring and diagnosing a system having logical reasoner applications, signal handler applications and signal broker module each as separate entities which can function on separate and distinct computers. The use of separate and distinct networked applications serves to enhance the scalability of the diagnosis server and the overall performance thereof as well as the convenience and practicality of using a diagnosis server networked for use by multiple users in multiple remote locations using multiple computers.
Another advantage of the invention is that it provides a diagnosis server for remote continuous monitoring and transmittal of sensor generated signals, to logical reasoner applications which process the signals and provide analysis, diagnosis, detection and isolation of system component failures and maintenance of systems of various complexity in real-time from a network remote to the system.
Thus, the present invention real-time diagnosis server and method promotes safety and efficiency in the operation of systems by providing continuous and accurate monitoring, diagnosing, and maintaining of complex systems. Furthermore, the invention reduces the operation costs of complex systems by identifying failures as they happen or potential failures prior to happening which can greatly reduce the time and expense of manual diagnosis and identification of system failures.
The above and other objects and advantages are satisfied by the invention which is a networked-based diagnosis server and method for monitoring and diagnosing a system having multiple components. The diagnosis server being remote from the system being observed.
The network-based diagnosis server comprises at least one sensor for monitoring and generating signals indicative of at least one characteristic of a component of the system, at least one sensor agent coupled to the sensor for receiving signals or messages therefrom, an interface connecting the sensor agent to the network, and a broker module coupled to the network for sending messages to and receiving messages from the sensor agent. The diagnosis server further comprises at least one handler application initiated by and connected to the broker module for transmitting messages to and receiving messages from the broker module and at least one reasoner application in communication with and invocable by the handler application for processing, and responding to messages received from the handler application. The handler application can include observer applications for monitoring the signals being sent and received by the broker module.
The following applications: sensor agents, broker module, handler and reasoner applications operate simultaneously relative to each other providing real-time monitoring and diagnosing of the system being observed. The broker module is independent of the handler applications and reasoner applications such that the diagnosis server is readily adaptable to various systems. Thus, the diagnosis reasoner of the present invention provides continuous monitoring and diagnosing of multiple components of the system in real-time.
The preferred embodiment of the invention comprises at least one reasoner application having a user interface for user controlled processing and responding to messages received from the handler application. The user interface allows the user to update the database as necessary with current information for the particular system.
Alternatively, the diagnosis server can maintain a system in conjunction with the monitoring thereof wherein the component of the system includes a controller coupled directly to the signal agent of the diagnosis server for receiving signals generated by the supervisor reasoner for controlling the component of the system. The control signals are transmitted to the controller via the signal agent.
The present invention diagnosis server provides real-time monitoring and diagnosing of the system it is observing by continuously processing and responding to periodic messages transmitted to the signal agent of the diagnosis server by the sensors which monitor the components of the system and generate test results or signals indicative of at least one characteristic of a component the system. Typically the sensors may generate and transmit periodic signals in the range of every one to two seconds. The signals transmitted by the sensors are subsequently received, processed and responded to by the diagnosis server such that the system is monitored, diagnosed and maintains the system continuously and in real-time. The invention provides a diagnostic server that is built on a three-tier architecture having a client layer consisting of the sensors and sensor agent applications. The sensor agent applications receive signals from the sensors, condition the signals and transmit them over a network to the broker module.
The middle layer consists of the broker and handler applications for transmittal of the sensor signals to the appropriate reasoner applications. The broker also performs session management, flow control, management of shared memory, message buffering and routing, and computational load balancing.
The third layer, the xe2x80x9cback-endxe2x80x9d consists of multiple on-line and user-interactive reasoner applications and accessible databases for processing signals received from the handler applications, diagnosing the status of the components of the system as well as detecting failures in the system.
The diagnosis server of the present invention functions on computers networked through the Internet, Local Area Network (LAN), or other Transmission Control Protocol/Internet Protocol (TCP/IP) network.
The three-tier architecture provides for efficient monitoring of systems of various complexity, such that a 50 MHz processor can support tens of concurrent sessions involving hundreds of tests. The solution scales to hundreds of sessions in any modern workstation or server.
The above stated performance results for the present invention diagnosis server are derived from the applicant""s testing of the broker module, an RT-Agent handler application and a TEAMS-RT logical reasoner application on a Sun SS20/502 system with 50 Mhz Super SPARC processor and 224 Mb of RAM. (The RT-Agent handler application and TEAMS-RT logical reasoner application are identified below.) A slow computer was used for testing purposes for the diagnosis server because the timer routine results were unreliable in the sub-ms range such that the run time for the diagnosis server could not be accurately measured using the faster processors. The sensor agent client applications were run on both Sun and Windows NT PCs with various hardware configurations. A web-server was also used to serve the monitoring console on the Sun SS20 system. The test results shown in Table 1 following, confirm that using the distributed architecture disclosed herein, the SS20/502 can provide remote diagnosis capability to hundreds of clients, as the reasoners are run on faster remote computers. Results from other tests of the present invention diagnosis server are available in the papers entitled xe2x80x9cTele-Diagnosis: Remote monitoring of large-scale systemsxe2x80x9d Mar. 15, 2000, Deb, Ghoshal, Malepati, Kleinman, all of Qualtech Systems, Inc., 100 Great Meadow Road, Suite 501, Wethersfield, Conn. 06109 and xe2x80x9cRemote Diagnosis Serverxe2x80x9d, Oct. 7, 2000, Deb, Ghoshal, Malepati, Cavanaugh, all of Qualtech Systems, Inc., 100 Great Meadow Road, Suite 501, Wethersfield, Conn. 06109. The above-identified articles are incorporated herein by reference.
The preferred embodiments of the invention utilize prior art TEAMS-RT logical reasoners developed by Qualtech Systems, Incorporated, located at 100 Great Meadow Road, Suite 501, Wethersfield, Conn. 06109. The TEAMS-RT reasoner application is described in detail in xe2x80x9cDe-centralized Real-time Monitoring and Diagnosisxe2x80x9d, Deb, S., Mathur, A., Willet, P., and Pattipati, K. R., the paper is available at the Qualtech Systems, Incorporated website and incorporated herein by reference.
The key features of the TEAMS-RT reasoner are the following: 1) separation of system specific information, represented in terms of models, from the fault-isolation methods. This allows for the same reasoner application to be used on multiple systems using different models; 2) ability to diagnose multiple failures in fault-tolerant systems with multiple modes of operation; and 3) excellent performance using low-end microprocessors.
The TEAMS-RT reasoner is therefore ideal for real-time diagnosing and monitoring of various systems. It is currently used for monitoring systems as diverse as helicopter engines and transmissions to bus systems in the International Space Station. However, embedding additional software in existing systems can introduce safety and validation concerns, especially in applications involving aircraft or the aerospace industry. Therefore the present invention diagnosis server designed to operate remote from the system being monitored is especially advantageous for these and many other applications.
The preferred embodiment of the invention further comprises prior art TEAMATE, user-interactive reasoners also developed by Qualtech Systems, Incorporated. The TEAMATE reasoner is based on Qualtech Systems, Incorporated, TEAMATE tool, which is an adaptive, intelligent diagnostic engine for field (offline) maintenance. The TEAMATE reasoner application provides network based adaptive and interactive diagnosis and is described in detail in the article xe2x80x9cAn Integrated Process for System Maintenance, Fault Diagnosis and Supportxe2x80x9d, Ghoshal, S., Shrestha, R, Ghoshal, A., Malepati, V., Deb, S., and Pattipati, K., also available at www.teamqsi.com/rds and incorporated herein by reference.
The TEAMATE reasoner as used with the present invention is of great value to a field technician as it allows him to conduct the diagnosis on a local system and access through the diagnosis server information and models that fit the specific system being diagnosed. The diagnosis server allows for data and Interactive Electronic Technical Manuals (IETMs) to be readily accessible and kept safe and secure as opposed to hard copy information and manuals.
The preferred embodiment of the invention also includes the TEAMS-KB diagnostic database. The TEAMS-KB diagnostic database is a product of Qualtech Systems, Incorporated, (QSI). TEAMS-KB stands for Testability Engineering and Maintenance System-Knowledge Base and is a companion tool to other QSI tools. TEAMS-KB is a member of QSI""s Integrated Tool Set and serves as a model management and maintenance data collection tool that can be deployed at various maintenance sites. It is used for model management, scheduled and unscheduled maintenance, diagnostics data collection, statistical data analysis, and data mining for trend and anomaly detection/isolation.
Features of TEAMS-KB include the management of diagnostic models and the service of diagnostic and health status information for onboard and ground-based fault isolation and troubleshooting tools. The efficiency of the reasoning engines behind these tools depend on parameters such as the failure rate of components, costs and time required for performing diagnostic tests, error likelihood of test outcomes, and accurate dependency information relating faults to failures, symptoms, and anomalies. The updating of such parameters using mined output of historical data collected from various fleets would greatly contribute to efficient maintenance by reducing time to troubleshoot and repair, and increasing availability.
TEAMS-KB can collect data automatically from TEAMATE, which eliminates the delays and errors of manual data collection. The on-line diagnostic and maintenance information collected by TEAMATE (components repaired, repair times/costs, test times/costs, etc.) can be archived by TEAMS-KB. TEAMS-KB provides an important link to the integrated diagnostic process. TEAMS-KB includes various parameter estimation algorithms for the analysis of maintenance data. It can be used to refine a system model residing in the knowledge base of TEAMS by updating repair costs, repair times, component failure rates and diagnostic costs and times. Additionally, TEAMS-KB can be utilized efficiently at maintenance sites as the analysis algorithms included can provide predictions of optimal shelf-stocks of spares and parts requirements as well as other information. The reports generated by TEAMS-KB can provide insight into operations, which slow down the entire maintenance process by pinpointing unreasonably long task times, repeated failures of certain components, and updating hazard rates of components caused by changing environmental conditions.
TEAMS-KB can automate scheduled and non-scheduled maintenance on a wide range of systems including manufacturing equipment and vehicles. TEAMS-KB can also be used to generate work orders, track maintenance and parts history, and create maintenance reports. Using TEAMS-KB, one can quickly review maintenance schedules or automatically generate work orders therefor. Additionally TEAMS-KB provides a means to efficiently track numerous business and maintenance related costs and activities such as equipment or facility repair or maintenance costs, labor costs, equipment history or performance records, employee performance records, work orders, and repair times.
The TEAMS-KB diagnostic database provides model management and diagnostic data management as its principal features. Model management includes the capability to create and modify models and components that comprise the model. The TEAMS-KB diagnostic database application can manage models, test models and provide an integrated environment for model development. Diagnostic data management includes capturing and managing diagnostic test and session history logs, managing schedules, and tracking and predicting spare part requirements.